Dr. Alexander Kashlinsky is a principal investigator on several NASA and NSF grants studying topics related to cosmological bulk flows, cosmic microwave and infrared background radiation, and early stellar populations. Using the Wilkinson Microwave Anisotropy Probe, Kashlinsky recently discovered a phenomenon called “dark flow,” which are clusters of galaxies moving at a constant velocity toward a 20-degree patch of sky between the constellations of Centaurus and Vella.

NASA Tech Briefs: You have a PhD in astrophysics from Cambridge University in England and your area of expertise at NASA is observational cosmology. What prompted you to pursue a career in this field?

Dr. Alexander Kashlinsky: It actually started in my youth from reading too much science fiction, which I no longer do. I distinctly remember how it was triggered. I picked up a book from the shelf by Stanislaw Lem, called the “Magellanic Cloud,” which was about the first interstellar travel, and it conquered my mind at the time, but I ended up in astrophysics and not traveling to the stars. Later, when I was doing my PhD, I was very privileged to work with Martin Rees, who is a very inspirational scientist, and that triggered my interest in astronomy and particularly in cosmology. He was very open-minded, and very interested in completely different ideas, which I found very stimulating and very inspiring. The rest is history.

NTB: Several years ago you were part of a team that succeeded in isolating the energy radiated by the first stars formed after the Big Bang, called Population 3, from all other energy that makes up the cosmic infrared background. What did you learn from that breakthrough?

Kashlinsky: What we did at the time was, we analyzed very deep data available thanks to the Spitzer Infrared Telescope, and we were trying to find how much diffuse radiation is left after we removed the various contributions that we can isolate in the images. What we learned is that the residual diffuse background – the so-called cosmic infrared background radiation – has quite a bit of energy emitted from sources that are much too faint to be detected, even in deep Spitzer exposures. That most likely means that these sources are very far away, because we removed galaxies down to a very faint level, that is, very far away. They had very little time to radiate all this very substantial energy that we detected and, therefore, they had to be quite abundant and they had to be radiating at enormous rates compared to typical populations living today. That, in our opinion, meant these populations were dominated by very massive stars, or very massive black holes, that lived very short times, but each unit of their mass emitted so much more energy than present day stars – such as the sun – that they had to produce this signature.

What is important in this context is not only what we learned, but what we did not learn. With the current data we could not learn whether these sources were stars that emit their energy by converting hydrogen into helium, or they were massive black holes that existed in very early times and that emitted energy by accretion processes, by gas falling into them and emitting energy in the process.

NTB: More recently, using NASA’s Wilkinson Microwave Anisotropy Probe, you discovered a phenomenon you refer to as “dark flow.” What is dark flow?

Kashlinsky: What we set out to measure in that measurement was the so-called peculiar velocities of clusters of galaxies, which are deviations from the uniform expansion of the universe. We never expected to find what we found at the end. We designed a method several years ago to probe the expected – within the standard cosmological models – peculiar velocities. The trick was to use many, many clusters of galaxies whereby you detect a very faint signal by beating down the noise. So we teamed up with colleagues at the University of Hawaii who assembled this x-ray cluster catalogue, and applied that method to the Wilkinson Microwave Anisotropy Probe, and we were very surprised by the results.

We found a flow that does not decrease with distance as far as we could tell, and we could probe to several billion light years away from us. It roughly was a constant amplitude, whereas in the standard cosmological model you expect that it should’ve been decreasing linearly with increasing scale. That is, as you go from, say, a few hundred-million light years to a few billion light years, it should decrease by an order of magnitude. We did not find that. We found a more or less constant velocity all the way as far as we can probe. The reason we called it dark flow is because the matter distribution in the observed universe, which is very well-known from galaxy surveys and from cosmic microwave background anisotropy measurements, that matter distribution cannot account for this motion. So this is why we suggested that if this motion already extends so far, then it probably goes all the way across the observable universe to the so-called cosmological horizon, and it is caused by the matter inhomogenity, or, I should say, space-time inhomogenity, at very large distances well beyond the cosmological horizon, which is about 40 billion light years away from us.

NTB: It’s been theorized that this dark flow may somehow be related to inflation, the brief hyper-expansion of the universe that occurred shortly after the Big Bang. Can you explain that relationship to us?

Kashlinsky: Yes. What we suspect is happening is the following. Inflation was designed, if I’m not mistaken in the early 1980s, to explain why the universe we see around us is homogeneous and isotropic. It is homogeneous – it’s roughly the same on all scales – and isotropic – it’s roughly the same in every direction.

Now, the way inflation works is as follows. It says that at some very early time the universe, or the underlying space-time, was not homogeneous. What happened then was that there was some bubble, a very tiny bubble of space-time, which, by pure chance, happened to be homogeneous by purely casual process, and then because of the various high-energy processes in the early universe this bubble, along with the rest of the space-time, expanded by a huge amount. We, today, live inside a tiny part of that original homogeneous bubble and we, therefore, see the universe around us as homogeneous and isotropic because the scales of inhomogenities that are other bubbles have been pushed away very, very far. What it means, at the same time, is that the original space-time was not homogeneous. If we go sufficiently far away, we should see the remnants of the pre-inflationary structure of the universe, of the space-time. These remnants would cause a very long wavelength wave across our universe and because there would be a gradient in this wave from one edge of the universe to the other, or from one edge of the cosmological horizon to the other, we would see a certain tilt, or the matter would be flowing from one edge to the other.

The analogy I could think of is as follows. Suppose you are in the middle of a very quiet ocean and you see the horizon, which determines how far you can see. As far as you can see, the ocean is isotropic and homogeneous. You would then think, at first, that the entire universe is just like what you see locally, that it’s homogeneous and isotropic like your own horizon. But then, inside that ocean, you discover a very faint stream from one edge of the horizon to the other...a flow. From the existence of that flow, you could deduce that somewhere very far away there should be structures that are very different than what you see locally. There should be mountains for this flow to flow from, or some ravines for this flow to fall into. So that would give you a probe of what the underlying very large scale structure of what your universe, or space-time, or some today call it multiverse, is that it is not just like what you see locally, but that sufficiently far away your space-time is very different from what you see here. So, in that sense it’s very much in agreement with the underlying inflationary paradigm that the initial space-time was very inhomogeneous, and we just happen to live inside a very homogeneous and isotropic bubble, but if we were to go very, very far away, we should be able to see such inhomogenities.

NTB: The galaxy clusters that make up this dark flow are rapidly moving toward a 20-degree patch of sky between the constellations of Centaurus and Vella. Why there, and do we know what’s attracting them?

Kashlinsky: Our limit on the 20-degree patch is purely due to observational error. If we were to make this measurement with, say, an infinite sample of clusters of galaxies and infinitely noiseless cosmic microwave background data, we presumably would measure just one uniform direction measurement. Why there? It’s by pure chance. It just happens to flow in that particular direction. As for what’s attracting them, we know that such flow cannot be generated by the matter distribution inside the observable universe, inside the universe that we observe. So, we therefore concluded that it must be something else very, very far away from us that is attracting them.

NTB: What impact, if any, does the discovery of dark flow have on our understanding of the universe and how it works?

Kashlinsky: What it tells us is that what we call, today, the universe is part of the overall cosmos, the overall space-time, whose structure is very different than what we see locally. Today, various issues of terminology that, at first, what people would call universes essentially...people would think that this is all the space-time there is and the universe, by definition, is all that there is in it. Today, people start talking in terms of multiverse, and multiverse is then composed of the various universes such as our own — that is, our own cosmological horizon, or our own bubble in the terms of this inflationary language. But there could be various other universes in this multiverse, in this landscape in which we live.

So, in that sense, what these measurements may imply is that our universe is just one of many and others may be very different from ours, and that there is an underlying multiverse in which these universes exist. So, if you would, it could imply an ultimate Copernican principle. It could generalize it, ultimately, that not only is our planetary system one of many, and our planet one of many, our universe may be just one of many.

NTB: One of the projects you’re currently working on at NASA is called “Studying Fluctuations in the Far IR Cosmic Infrared Background with COBE FIRAS Maps.” Tell us about that project and what you hope to accomplish with it.

Kashlinsky: This project and the group of us working on it — it’s myself, Dave Fixsen, and John Mather here at Goddard Space Flight Center — what it is designed to measure is the structure of the cosmic infrared background radiation at far infrared bands.

Why COBE FIRAS? FIRAS is the Far Infrared Absolute Spectrophotometer that was launched onboard the COBE satellite, the satellite that discovered cosmic microwave background structure. That instrument measured the spectrum of the cosmic microwave background radiation and it determined that it is a basic black body spectrum, basically down to almost one part in 1,000,000. But it also gave us very useful maps to work with for other parts of science. It measured all sky maps at the various far infrared wavelengths. Because we know the spectrum of the cosmic microwave background radiation, we can remove it from these maps very well. Then, if we’re lucky — and by lucky I mean if we can remove other foreground, such as our own galaxy, sufficiently well — we can then determine how much is produced by distant galaxies at far infrared wavelengths. That will give us very important cosmological information as to how these galaxies lived when they produced these emissions, how much of these emissions they produced, and so on and so forth. This project is just beginning, so I don’t know what our results will be, but the hope is to isolate the fluctuations in the cosmic infrared background radiation after subtracting the cosmic microwave background radiation from the FIRAS maps.

NTB: You mentioned that one of your co-investigators on this project is Dr. John Mather, the 2006 Nobel Laureate in physics. Do you ever find yourself dreaming of one day possibly winning a Nobel Prize, or is that something scientists don’t really think about until it happens?

Kashlinsky: Oh, I think it’s the latter. It just doesn’t cross the mind, I would say, of most scientists because you are so busy trying to understand whether the results you are measuring are real; what the systematics are; what the statistical significance is; whether you have been fooled by the various other processes that you have not accounted for; that it doesn’t give you much time to stress or share thoughts. So no, I don’t spend time thinking about it. And once you produce results, you really are worried that these are real results; that they can be maintained by future measurements; and you should always seek confirmation of these results, so no, there’s not much time to think about that.

NTB: What are some of the other significant projects you’re either working on, or anticipate working on, in the future?

Kashlinsky: It’s a very fortunate era now in the field of cosmology. I remember when I was starting my PhD, there was very little data to go by and there were many ideas, but also the theoretical part of the field was not particularly developed as I look back at it now. Slowly but surely, theoretical understanding developed and then, what’s even more important, in the last, I would say, ten or fifteen years there has been an explosion in the data – high quality data – that has been obtained in this field. This data comes from various space observatories or satellites such as the COBE satellite. It was a very important point in cosmology, and it was reached also thanks to the new generations of ground telescopes that can see very far with very high resolution and very low noise.

So, today you have a lot of data that can really constrain your understanding of the theoretical issues of the universe, and these data come at various wavelengths. For instance, in terms of cosmic microwave background measurements, there was COBE, then there was WMAP (Wilkinson Microwave Anisotropy Probe), which is still operating. It’s a superb instrument. And the Europeans are going to launch, this spring, the successor to WMAP, called the Planck satellite, which should bring a lot of new cosmic microwave background radiation data over a very wide range of frequencies with very low noise and with fairly good angular resolution. That is one of the projects we’re thinking to do with the dark flow studies; we want to try it with the Planck data.

At other wavelengths there is the Spitzer satellite, which is still operating. It is now about to begin its so-called warm mission because it’s run out of cryogens, so it has been extended for warm mission and it should still bring some very important data for understanding distant populations and the cosmic infrared background radiation emitted by them.

You can also go to a completely unexpected range of wavelengths or energies. At very high energies there is now operating… The GLAST (now renamed Fermi) satellite, which is the successor to the Compton Gamma Ray Observatory, and it is going to map the universe very well at gamma ray wavelengths and find a lot of distant gamma ray sources, gamma ray bursts, and so on. This would also be important in terms of studying early stellar populations because – this is one of the projects I hope to do with the data – you should see a very distinct cutoff in the spectrum of gamma-ray sources (bursts and blazars) at very large distances. This is produced by the cosmic infrared background from very early sources, such as Population 3, or the first black holes, and it is produced because the energy that these sources emit, which reach us in the infrared band, would also contain a lot of photons, and the very high-energy photons produced by these gamma ray bursts then would flow in the sea of IR photons – the cosmic infrared photons produced by the first stars – and they would get absorbed at sufficiently high energies by the so-called photon-photon absorption process. So, you should see a certain spectral feature that would tell you, yes, this is the epoch where these first stars lived. Maybe they lived for the first hundred-million years, maybe they lived for the first two-hundred-million years, and so on. You should be able to see the feature, if they produced enough energy.

And, of course, there are preparations for science that can be done with the James Webb Space Telescope, the JWST, which is going to be launched four or five years from now. That would be a successor to Hubble, but it also measures the universe in infrared bands, so it would see very far. It would see at completely different wavelengths, and it would bring a lot of data and probably revolutionize our understanding of the evolution of the universe.

For more information, contact Dr. Alexander (Sasha) Kashlinsky at This email address is being protected from spambots. You need JavaScript enabled to view it..

To listen to this interview as a podcast, click here